Neuropeptide Technology: Gene Expression and Neuropeptide Receptors

volumes.

Methods in Neurosciences

Edited by

P. Michael Conn

Volume 1 Gene Probes

Volume 2 Cell Culture

Volume 3 Quantitative and Qualitative Microscopy

Volume 4 Electrophysiology and Microinjection

Volume 5 Neuropeptide Technology: Gene Expression and Neuropeptide Receptors

Volume 6 Neuropeptide Technology: Synthesis, Assay, Purification, and Processing (in preparation)

Volume 7 Lesions and Transplantation (in preparation)

Section I: Neuropeptide Gene Expression

Outline

Assays for Peptide Products of Somatostatin Gene Expression

Assay of Atrial Natriuretic Factor Messenger Ribonucleic Acid

Molecular Assays for Rat Thyrotropin-Releasing Hormone Gene

Assays for Corticotropin-Releasing Hormone and Vasopressin Messenger RNAs

Motilin: Structure, Expression, and Chromosomal Localization of Human Gene

Progonadotropin-Releasing Hormone Synthesis and Processing: Measurements of mRNA and Peptides

Use of Firefly Luciferase Reporter Gene to Study Angiotensinogen Acute Phase Response Element

Cell-Specific Localization of Neuropeptide Gene Expression: Gastrin-Releasing Peptide or Mammalian Bombesin

Semiquantitative Analysis of Cellular Somatostatin mRNA Levels by in Situ Hybridization Histochemistry

Measurement of Oxytocin and Vasopressin Gene Expression by in Situ Hybridization

1

Assays for Peptide Products of Somatostatin Gene Expression

J.J. Holst and M. Bersani

Publisher Summary

This chapter presents experimental results undertaken to study somatostatin gene expression in peptide products. Somatostatin belongs to the group of biologically active peptides that is usually called regulatory peptides. Somatostatin is of particular interest because it functions in the neuroendocrine regulation of hypothalamic secretion as a circulating hormone, as a neurotransmitter, and as a transmitter of paracrine regulation. The chapter describes the use of radioimmunological methods. The carbodiimide method is used to form peptide bonds between carboxyl groups and primary amino groups belonging to both the hapten and the carrier. An antigen does not have to be larger than about 10 amino acids. The preparation of a radioactively labeled peptide depends on the chemical nature of the peptide. The amount of charcoal required for complete separation of labeled and unlabeled peptides depends on the specific peptide and the specific assay. The incubation volumes, incubation times, incubation temperature, and the dilution of the reactants depend on the required sensitivity and the affinity of the antibodies. By choosing increasingly selective elution conditions and using combined monitoring of the UV absorption and radioimmunoreactivity, one may gradually increase the purity of the immunoreactive substance and eventually reach a sufficient purity for amino acid, mass, or sequence analysis.

Introduction

Somatostatin belongs to the group of biologically active peptides that is usually called regulatory peptides. This designation was made when it was realized that biologically active peptides are involved in endocrine, paracrine, and neural regulation of cell and tissue functions. Somatostatin is of particular interest because it functions in the neuroendocrine regulation of hypothalamic secretion (1), as a circulating hormone (2), as a neurotransmitter (3), and as a transmitter of paracrine regulation (4). Accordingly, somatostatin is produced in neurons and in paracrine and endocrine cells in many different tissues in the body, including the central nervous system (5), endocrine glands like the pancreatic islets and thyroid gland (1), and intrinsic neurons (3) and epithelial cells of the gastrointestinal tract (6).

In spite of the diversity in functions and occurrence of somatostatin, somatostatin is believed to be encoded in mammals by a single gene (7–9). The gene encodes a peptide, preprosomatostatin, of 116 amino acids (Fig. 1). The N-terminal 24 amino acid fragment appears to function as the signal peptide and is cleaved from the preprohormone efficiently and rapidly after initiation of synthesis (10), leaving behind the prohormone, prosomatostatin (Fig. 1). Subsequent proteolysis cleaves the prohormone into smaller fragments, and among these is the 14-amino acid C-terminal fragment, somatostatin. However, it is now clearly established that the different tissues produce additional molecular forms which contain the somatostatin sequence, and these molecular forms occur in amounts which differ markedly between tissues. Because differential processing of the proRNA encoding somatostatin has not been identified in mammalian tissues (8,9), that is, only a single mRNA species can be identified regardless of the tissue of origin, it follows that the diversity of forms is a consequence of a tissue-specific, differential processing of the prohormone. Because there is, as yet, no method that can predict the pattern of posttranslational processing of peptide precursors in the different tissues, it is necessary to extract and to analyze the products actually occurring in each tissue in order to identify the chemical nature of the regulatory peptides produced and to study the physiological significance of their production/release.

Fig. 1 Diagrammatic representation of preprosomatostatin and its processing products, signal peptide and prosomatostatin. Prosomatostatin is further and differentially processed in ileal mucosa and pancreas, as described in the text. In ileum the products are proSS 1–64 and proSS 65–92, whereas in pancreas they are proSS 1–64, proSS 65–76, and proSS 79–92.

The major products of prosomatostatin (proSS) occurring in the pancreatic islets are the following (Fig. 1): somatostatin 14 (proSS 79–92) (11), somatostatin 28 (1–12) (proSS 65–76) (12), and prosomatostatin 1–64 (13). The major products of prosomatostatin that occur in the mucosa of the small intestine (Fig. 1) are somatostatin 28 (proSS 65–92) (14, 15) and prosomatostatin 1–64 (13). Other molecular forms identified so far include proSS 1–32, isolated from small intestine (16) (but this form may represent an extraction artifact); proSS 1–10, isolated from extracts of the antral mucosa (17) and also designated antrin; and, possibly, the entire prosomatostatin molecule (18). Tumors or tumor cell lines that express the somatostatin gene may produce a multitude of molecular forms (19–21), and a large number of molecular forms occur in extracts of the central nervous system (22), among these the already described forms proSS 65–76, proSS 65–92, and proSS 79–92.

Methods for Detection of Products of Prosomatostatin

In theory, almost any fragment of proSS might be formed in cells that express the SS gene; therefore, an array of analyses that are directed against very small sequences of proSS are required for a full analysis of the processing. Because the products may occur in very small amounts (e.g., in peripheral nervous tissue), the analyses must also be very sensitive. This generally means that the most suitable method will be a radioimmunological analysis or related technique with a similar specificity and sensitivity. Only radioimmunological methods are discussed in this article.

Antiserum Production

The antiserum binding site will accommodate between 4 and 8 amino acids, which means that the specificity of the antiserum can be expected to correspond to sequences of a similar length. To allow detection of such a sequence in a molecule that is larger, that is, extended in either or both of the terminals of the fragment, it is necessary that the antiserum not be directed against these termini. In other situations, however, one may wish to identify by radioimmunoassay the exact chemical structure of a peptide terminus, and this can be achieved if the antibody exclusively binds the unextended and unmodified C terminus of the peptide. For the production of antibodies, therefore, it is desirable to produce an immunogen that exposes the exact structure to be measured. This can be achieved by selecting appropriate methods for covalent coupling of peptide fragments to a suitable carrier protein as described below.

Although the actual processing of the precursor peptide cannot be predicted, certain rules have emerged. Many propeptides are cleaved at sites where two basic amino acids occur (10); cleavage may also occur at the site of a single basic amino acid residue (10). A large proportion of the naturally occurring biologically active peptides are amidated at the carboxy terminus by a process involving cleavage of basic residues by the carboxypeptidase B-like processing enzyme and transfer of the nitrogen from a glycine residue to the preceding amino acid, a step catalyzed by the amidating enzyme (23). If a glycine residue occurs N-terminally to a suspected basic cleavage site, an α-carboxyamidation may also occur. With this background one can design the immunogen for antibody production.

Synthesis of Haptenic Antigen

As already mentioned, the antigen does not have to be larger than about 10 amino acids. Peptides of this size are relatively easily produced by chemical synthesis (24, 25). Automated equipment is available (26), and an amide group can also be introduced. In addition, most of the chemical companies that produce peptides also offer custom synthesis of peptides. Whether produced in the laboratory or purchased from a manufacturer, the fragment may need purification as described below.

Preparation of Immunogen

Procedures for immunogen preparation may be found in many handbooks on radioimmunoassay technology (27). The following methods have been successfully used in our laboratory for production of antisera against somatostatin and other small peptides: (1) conjugation with carbodiimides, (2) conjugation with glutaraldehyde, (3) conjugation with difluorodinitrobenzene, and (4) conjugation via cysteine thiol groups.

Carbodiimides

The carbodiimide method is a modification of the method originally described by Goodfriend et al. (-amino groups, then coupling is likely to proceed through the amino terminus. In theory, the carbodiimide is thought to activate carboxyl groups first, which will subsequently react with amino groups (27). If the carbodiimide is briefly preincubated with only one of the reagents before addition of the other, the coupling may be directed accordingly. This method, however, will merely enhance the probability of a certain direction of coupling rather than producing a homogeneous product.

The procedure that we use is designed for small amounts of peptide, namely, between 0.5 and 2.0 mg, which will frequently be the maximum amount of a peptide available when purchased from a chemical company.

1. Dissolve 0.5 to 2.0 mg of peptide in a small volume (100 μl) of solution, the pH of which is sufficiently different from the presumed isoelectric pH of the peptide to ensure that the peptide dissolves (e.g., dilute hydrochloric acid). Add sodium phosphate buffer (50 mmol/liter) so that the total volume corresponds to the number of milligrams of peptide (0.5 to 2.0 ml).

2. Add 2–5 mg of bovine serum albumin (Cohn fraction V from Sigma, St. Louis, MO, No. A-4503).

3. Dissolve 50 mg of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (No. E-7750, Sigma) in 10 ml distilled water. Add 200–350 μl dropwise to the peptide-albumin solution with gentle stirring.

4. Wrap the tube containing the mixture with aluminum foil (or put in a dark room) and mix gently (with a magnetic stirring bar) overnight.

5. Dilute the mixture with 0.15 mol/liter NaCl sufficient for the number of animals that one expects to immunize (see below) and divide into portions corresponding to each immunization. Note that it is neither necessary nor advisable to purify the conjugate. Purification, however, may easily be accomplished by gel filtration (see below).

Glutaraldehyde

The glutaraldehyde method was originally described by Avrameas (29). We use the following modification for the production of antibodies against the 1–12 fragment of somatostatin 28 (30).

1. Dissolve 0.5 mg peptide in 100 μl of 1 mmol/liter HCl.

2. Add 0.9 ml of 50 mmol/liter sodium phosphate buffer, pH 7.4.

3. Add 2.5 mg bovine serum albumin as above.

4. Add dropwise 200 μl of a 12.5% (v/v) solution of glutaraldehyde. Stir with a magnetic stirrer, wrap the tube in aluminium foil, and continue stirring for 3 hr at room temperature.

5. Purify the immunogen by passing it over a small (e.g., 0.9 × 10 cm) column packed with Sephadex G-10 (Pharmacia, Uppsala, Sweden) and elute with 50 mmol/liter phosphate buffer as above. The void volume is collected. (If one adds a trace amount of ¹²⁵I-labeled albumin to the incubation mixture, the void volume is easily identified by counting eluted fractions.)

6. The collected fraction(s) is diluted further with saline as described below and stored in appropriate portions at −20°C.

Difluorodinitrobenzene

The difluorodinitrobenzene method has been described in detail by Tager (31). The following procedure was successfully used for the production of C-terminal glucagon antibodies.

1. Ten milligrams of peptide is dissolved in 1 ml of 0.1 mol/liter phosphate buffer, pH 7.2, containing guanidine hydrochloride, 7 mol/liter.

2. Add 150 mg difluorodinitrobenzene (Sigma) dissolved in 5 ml freshly distilled methanol and mix well. After 15 min at room temperature, the reaction mixture is cooled on ice. The precipitate (the activated peptide) is washed thoroughly with chilled diethyl ether (using a cooling centrifuge safeguarded against explosions).

3. The precipitate is dissolved in 5 ml borate buffer, 0.4 mol/liter, pH 10, containing 20 mg albumin, and the resulting mixture is allowed to stand overnight in the dark at room temperature. The conjugate may be dialyzed against phosphate-buffered saline before storage.

Cysteine Thiol Groups

The cysteine thiol method, which we have used on several occasions for high-efficiency, N-terminal couplings of synthetic peptides to a carrier (typically keyhole limpet hemocyanin), is particularly useful for synthetic peptide fragments because they may be designed with cysteine residues at either or both terminals. The method is essentially that of Liu et al. (32), as modified by Dyrberg and Kofod (33).

1. Keyhole limpet hemocyanin (KLH) is dissolved in 10 mmol/liter phosphate buffer, pH 7.2, to a concentration of 15.6 mg/ml and dialyzed for 24 hr. Use 5 mg of KLH for 5 mg of peptide.

2. Dissolve m -maleimidobenzoyl-N -hydroxysuccinimide ester (MBS) in dimethylformamide to a concentration of 12 mg/ml immediately before use.

3. Slowly add 55 μl MBS to 320 μl KLH (representing 5 mg) and incubate for 30 min at room temperature.

4. Isolate the activated KLH by applying the mixture to a 20- to 25-ml column of Sephadex G-25 equilibrated and eluted with 50 mmol/liter phosphate buffer, pH 6.0. Measure the effluent absorbance at 280 nm; the first peak to elute is the activated KLH.

5. Dissolve 5 mg peptide in 1 ml degassed redistilled water. Add 5 mg of activated KLH (assume 100% recovery in Step 4) to the sample under constant agitation. Adjust the pH to 7.0–7.5 (with NaOH or HCl) and incubate under constant agitation for 3 hr. Store the conjugate at −20°C appropriately diluted.

Immunization

For immunization we use young White Danish rabbits. We usually immunize four animals at a time. The general scheme followed is as follows: the animals are immunized 5 times at intervals of 15 days and then another 2–3 times at intervals of 1–2 months. The animals are bled 10 days after the last 2–3 boostings. This means that each animal will receive 7–8 immunizations. The procedure with four animals thus requires 28–32 doses of immunogen. Since we usually give each animal 0.5 ml immunogen at a time, this means that the final volume of immunogen (each of the final steps above) must be at least 14 ml, and that the portions should be 2 ml each.

In our laboratory we use Freund’s adjuvant (Statens Seruminstitut, Copenhagen, Denmark) for enhancement of the immune response. The immunogen is mixed with an equal volume of Freund’s adjuvant (complete for the first injection, incomplete for the subsequent injections) and thoroughly mixed until completely emulsified (this process is facilitated by using a Vibrofix VF-1, equipped with a tube holder, from Janke & Kunkel, D-7813 Staufen, Germany). For each animal 1.0 ml of the emulsion is injected intracutaneously at multiple sites on the back. Bleedings are performed through ear veins.

Animals that have not shown a clear-cut immune response after the series of immunizations are sacrificed. Animals with a positive response may be kept and boosted at longer intervals. We have maintained rabbits with an excellent antibody production for up to 10 years.

Preparation of Tracer

The preparation of a radioactively labeled peptide depends on the chemical nature of the peptide. If the peptide contains a tyrosine residue, the peptide may be easily labeled with ¹²⁵I using any of the mild oxidation methods. This may be considered when designing the assay, so that a fragment or sequence that contains a tyrosine residue is chosen. If the peptide contains no Tyr but a histidine residue, ¹²⁵I may be introduced into the imidazolium group of the His using stronger oxidation. If neither His nor Tyr is present, the peptide may be labeled using the Bolton-Hunter technique whereby a prelabeled phenyl residue is coupled to a primary amino group using a succinic anhydride condensation reaction. Since introduction of the labeled moiety often will occur at the α-amino group at the N terminus, this method may be used to direct the iodination so that the C terminus remains unmodified and available for binding of a C-terminally directed antibody.

Labeling of Tyrosine Residue

In our experience the best results when labeling Tyr are obtained with the so-called stoichiometric chloramine-T method (36). The chemical nature of the iodination reaction is not different from that obtained using other methods of oxidation [Iodogen, peroxidase (34)], but the extent of the incorporation of iodine may be completely controlled and the harmful effects of oxidation minimized. Our routine procedure is a modification of the method described by Roth (36).

1. It is convenient to label about 5 μg peptide. The vial used for the iodination reaction should be suitable for small volumes since the incorporation of ¹²⁵I is facilitated by high concentrations of reactants. The best procedure is to dispense a small volume of the peptide dissolved in a volatile buffer (e.g., acetic acid) into a number of tubes. Remove the solvent in a vacuum centrifuge or freeze dryer, close the tubes in a nitrogen atmosphere, and store at −20°C [Nunc Cryo tubes, No. 366656 (Teknunc A/S, Roskilde, Denmark), are suitable for this]. The peptide is dissolved in 5 μl of 0.3 mol/liter sodium phosphate buffer, pH 7.4 (this buffer is chosen to ensure neutralization of the NaOH present in the iodide solution and because it is a good solvent for some poorly soluble peptides).

2. ¹²⁵I is transferred to the solution. (Note : A well-ventilated hood with an iodide trap is required.) The iodide may be purchased as a solution with a radioactivity of 1 mCi/ml. For 5 μg of peptide we usually add 400 μCi (i.e., 4 μl).

3. A solution of chloramine-T (No. 2426, Merck, Darmstadt, Germany), 30 μg/ml, is prepared in the 0.3 mol/liter phosphate buffer. Small aliquots (e.g., 2.5 μl) of the chloramine-T solution are added to the peptide-iodide solution at intervals of at least 60 sec. With each addition an additional amount of ¹²⁵I is incorporated into the peptide. The number of additions depends on the degree of incorporation needed. This may be determined in preliminary experiments as described in detail by Roth (36). The principle is that a minute sample of the iodination mixture (e.g., what adheres to the point of a needle) is transferred to a 1% solution of albumin in buffer. An equal volume of a 20% (w/v) solution of trichloroacetic acid (TCA) is added, the mixture is centrifuged, and both the precipitate and the supernatant are counted for radioactivity. The ratio between the radioactivity present in the precipitate and that in the precipitate plus the supernatant represents the fraction which is incorporated. It is necessary to make sure that the peptide is precipitated by TCA (not all peptides are). For small peptides six additions of chloramine-T usually provide an incorporation of greater than 50% of the iodide.

The mixture is then diluted with a solution that is compatible with the subsequent purification method. For gel filtration it should contain a carrier protein such as 1% albumin (good quality, e.g., human serum albumin, Reinst Trocken, from Behringwercke, Marburg/Lahn, Germany).

Incorporation of ¹²⁵I into Histidine

We follow the procedure described by Schaffalitzky de Muckadell and Fahrenkrug (35).

1. To a reagent tube containing about 5 μg of peptide 5 μl of 0.3 mol/liter sodium phosphate, pH 7.4, is added.

2. Add 2 mCi of ¹²⁵I (often 20 μl).

3. Add 10 μl of a 4 mg/ml solution of chloramine-T in 0.3 mol/liter sodium phosphate buffer and stir for 60 sec (use a capillary pipette for the transfer of the iodide and for stirring; mixing can also be ensured by tapping the vial with a fingernail).

4. Add 10 μl of a 10 mg/ml solution of sodium metabisulfite (No. 6528, Merck) in phosphate buffer in order to stop the reaction.

5. Add a solution that is suitable for the subsequent purification of the peptide.

Iodination with Bolton-Hunter Reagent

We follow the recommendations of the manufacturer (Amersham International, Buckinghamshire, England) and use the special iodination vial supplied by them (No. IM 5861X, Amersham). The peptide in a suitable solvent is transferred to the iodination vial. As described above, the peptide may be stored in dry form on the bottom of a nitrogen-filled ampoule. In this case the Bolton-Hunter reagent solution may be transferred to the peptide-containing reagent tube and dried down by a stream of nitrogen (in a ventilated hood). The reaction may then be started by addition of an aqueous buffer as recommended by the manufacturer (usually borate buffer, pH 8.0). The reaction mixture is left on ice in a cold room overnight before further purification of the labeled peptide.

Purification of Labeled Peptide

Numerous purification methods have been described and used previously, but HPLC techniques are the most advantageous. We routinely purify labeled peptides using the following technique.

1. The HPLC column is a 0.4 × 25 cm Vydac C18 column (Separations Group, Ltd., Deeside, England) mounted with a Rheodyne valve equipped with a 1.0-ml loop for injection.

2. We generally use a linear gradient of acetonitrile (HPLC grade, No. 1015, Rathburn, Walkerburn, Scotland) in water containing 0.1% trifluoroacetic acid (TFA) (Pierce, Rockford, IL) that reaches 20 to 40% (v/v) acetonitrile in 100 min (or a gradient with a 0.2% increase in acetonitrile concentration per minute).

3. The iodination mixture is diluted with the TFA-containing water, and 500 μl is injected. Thorough washings of the injection valve, injection needle, and column before and after each run are essential for reproducible results.

By conducting radioimmunoassays of the effluent either in „dummy" experiments with incorporation of nonradioactive iodide or in control experiments with unlabeled peptide, the elution times of the unlabeled peptide may be compared to the elution times of labeled peptide; if the two are well separated, the specific activity of the labeled peptide can be assumed to be close to maximal, that is, up to 70 MBq/nmol depending on the isotopic abundance of the source of ¹²⁵I. Alternatively, the specific activity may be assessed from self-displacement experiments in the radioimmunoassay.

Usually, about 4–6 radioactive peaks emerge from the HPLC purification. Each of the peaks should be tested for binding activity because they may retain immunoreactivity. The elution pattern for a specific peptide is generally highly reproducible.

Separation System

With the small peptides usually employed for region-specific radioimmunoassays the separation technique is generally straightforward. However, the natural peptide to be identified may be much larger than the fragments used for development of the assays, and therefore special separation systems may be needed. An example from our laboratory is the assay of glicentin, the 69-amino acid N-terminal fragment of proglucagon. The only separation technique that permitted radioimmunological quantitation of this peptide was the double-antibody technique performed as described in detail elsewhere (37).

For routine separations we employ the plasma-coated charcoal technique. The amount of charcoal required for complete separation of labeled and unlabeled peptides depends on the specific peptide and the specific assay. As a first approach (which is most often successful) we employ the following procedures. Three grams of activated charcoal (No. 2186, Merck) is mixed with 250 ml of 50 mmol/liter sodium phosphate buffer, pH 7.4, containing in addition thiomersal, 0.6 mmol/liter. To this mixture is added 50 ml of plasma. The source of plasma is not critical. We have used outdated human plasma, horse serum, porcine plasma, and bovine plasma. (As a safety precaution, most laboratories should probably avoid human plasma.) The mixture is incubated in a cold room for at least 2 hr before use, and then 1.5 ml of this mixture is dispensed into the radioimmunoassay incubation mixture. This volume is suitable for an incubation volume of 500 μl. It is advisable to let the resulting mixture incubate for at least 30 min before centrifugation to prevent assay drift. The mixture is centrifuged at 4°C at 3000 rpm in a large radioimmunoassay centrifuge for 20 min or until the supernatant is completely clear. The resulting precipitate is firm, and decantation is easy. The basis for this separation technique is that any differences in protein content between samples and standards, and within samples, will be evened out by the addition of the relatively large amount of protein represented by the plasma.

Incubation Conditions

The incubation volumes, incubation times, incubation temperature, and the dilution of the reactants will depend on the required sensitivity and the affinity of the antibodies. The greater the required sensitivity, the larger the sample volume and the greater the dilution of the antibody solution. Preincubation of unknowns or standards with antibody before tracer addition may yield enhancement of sensitivity. When used as a monitoring method during isolation of natural peptides, a tabletop version of the radioimmunoassay may be employed. In such assays, performed at room temperature, more concentrated (2–3 times) reagents are used. Usually a reasonable equilibrium is reached within 1–3 hr, at which time separation may be performed. This procedure is usually sufficiently sensitive for the isolation of picomole amounts of neuropeptides from tissues and greatly facilitates the isolation procedure.

Extraction of Precursor Products from Tissue

The choice of tissue extraction procedure depends on the physicochemical characteristics of the peptide that one is determined to isolate. As discussed above, these may be deduced from the predicted precursor structure and the positions of predicted cleavage sites; however, there are numerous exceptions, and some general methods are therefore helpful.

During the course of a systematic survey of extraction methods for somatostatin in the pancreas a method was developed that is particularly suitable for isolation of neutral or basic peptides (11). Details of the method are presented below. Acidic peptides can also be extracted by this method, but some (e.g., the antral hormone gastrin and related peptides) are poorly recovered and should be extracted at a neutral pH as described below.

Method for Neutral or Basic Peptides

1. The tissue is processed while still frozen. Store the tissue in dry ice immediately before extraction. Weigh the tissue. Wrap it in a heavy plastic bag and crush finely with a hammer on a metal plate. The plate should be precooled with dry ice.

2. Add 4 volumes of acid-ethanol (660 ml of 96% (v/v) ethanol, 15 ml of 37% HC1, 125 ml distilled water) precooled to −20°C. The volume (ml) of the tissue is estimated as the weight (g) plus 10%. If very small amounts of tissue are extracted, the volume of acid-ethanol may be increased to ensure a satisfactory recovery. Such small tissue samples may then be homogenized in a Potter homogenizer. The larger samples are homogenized using a Waring blender of appropriate size. In both instances make sure that the temperature does not increase (cool with crushed ice).

3. Let the extract stand for at least 4 hr or overnight. Centrifuge at 4°C to obtain a clear supernatant and decant supernatant. (The precipitate may now be discarded or reextracted if a high recovery is essential.)

4. Add 5 volumes of high-quality diethyl ether (use newly opened bottles to minimize the risk of explosion and reduce the risk of oxidation) precooled to −20°C to the supernatant and mix well. Place the mixture on dry ice. A precipitate will form in the lower, aqueous phase. Decant the ether phase. (Note the danger of explosion. We always perform these steps out of doors.) Apply a stream of nitrogen (or compressed air) to the precipitate until the smell of ether can no longer be detected.

5. The precipitate can be reconstituted in distilled water. High-quality urea may be added to facilitate dissolution of the precipitated proteins. The extract is highly acidic and should immediately be subjected to further processing without attempts to reduce its volume further, because of the danger of acid hydrolysis. The virtue of this method lies in the fact that no protein which was initially soluble in the acid-ethanol has been discarded. The method will therefore ensure the most complete recovery of unknown substances.

Extraction of Small Samples and Acidic Peptides

Sometimes tissue samples for extraction are so small that the above procedure is not technically feasible. Such samples may include endoscopic biopsies or small animal tissues (e.g., the retina or pineal gland from rats). For such samples the following procedure may be used.

1. The tissue is weighed (if kept in a weighing tray positioned on dry ice, the tissue can be weighed while still in the frozen state) and submerged in 1 ml (or, for larger samples, 10 ml/g tissue) of boiling water (tubes preheated in an oil bath or an electric tube heating device; remember to cover to limit evaporation). Boil for 15 min.

2. Chill in ice. Homogenize in a Potter homogenizer. Avoid heating. Centrifuge in a high-speed centrifuge until the supernatant is clear. Decant and store the supernatant at −20°C until further processing. Note that this is a neutral extract that may contain acidic peptides.

3. Add 1 ml (or a volume equal to the initial water volume) of 1 mol/liter acetic acid to the precipitate. Resuspend and rehomogenize. Let stand for 1 hr in the cold, then centrifuge as above. Note that this extract (supernatant) may contain the remaining and possibly any neutral peptides.

4. The two supernatants can be combined or can be analyzed separately.

Further Processing of Tissue Extract

The choice for further processing of the tissue extracts depends on whether the purpose of the peptide isolation is mainly preparative or analytical (and this will usually determine the size of the extract). It is not advisable to perform the radioimmunoassay directly on the extracts because they may cause heavy nonspecific interference in most radioimmunoassays.

Analytical Processing

As discussed above the first and possibly most important posttranslational modification of a peptide precursor is proteolytic cleavage and the subsequent generation of fragments. In order to characterize the latter the most essential information is the molecular size, which is best investigated by gel filtration. For most peptide precursors (and certainly for prosomatostatin) the sizes of the resulting products range from a few to about 100 amino acids so that Sephadex G-25 to G-50 (Pharmacia) or similar gels will be suitable. To ensure that the equilibration medium for the gel filtration column is compatible with the extract to be applied, use acetic acid (0.5 up to 3 mol/liter, depending on the purpose of the gel filtration). The size of the column depends on the amount of extract being applied. The sample size should be about 2% of the column bed volume for optimal resolution. The column effluent is collected and the acetic acid removed by freeze drying or vacuum centrifugation. The residue may be reconstituted in assay buffer (and in this way even concentrated) and assayed directly. If the expected concentrations are high, the effluent may be assayed after a 1/50 dilution.

For characterization of eluted substances calculate the value of Kd, the coefficient of distribution (between mobile and stationary phases), as (Ve - Vo)/Vi, where Ve is the elution volume of the substance in question, Vo the totally excluded volume, determined as the elution volume of a large molecular weight marker (e.g., ¹²⁵I-labeled albumin), and Vi the available inner volume, determined as the difference between Vo and the elution volume of a small molecule (e.g., ²²NaCl). Determination of Kd in this manner is advantageous because the markers (¹²⁵I-labeled albumin and ²²NaCl in trace amounts) may be added to the sample to be filtered allowing an internal calibration of the column. In addition, the markers provide excellent quality control of the gel filtration; an unusual skew or broadness of the elution profiles of the markers will indicate that the resolution power of the columns is not optimal.

Eluted fractions may be pooled for further chromatographic analysis (in which case only a small sample of the effluent should be removed for radioim-munoanalysis). The acidic effluent pool may be applied directly to reversed-phase HPLC columns if mounted with sufficiently large injection loops or if the sample can be pumped onto the column from a special sample reservoir. It is advisable to include a precolumn in these systems. After the initial characterization of the gel filtration pool, the peptides may be eluted from the HPLC column with a gradient of acetonitrile. Any resulting immunoreactive peak may be further characterized in an isocratic HPLC system. By comparison with markers (synthetic peptides corresponding to the fragment suspected to be present in the extract) the eluted peptides may be identified as being identical to or different from the marker.

Preparative Processing

If large amounts of tissue are extracted for the purpose of isolating a new peptide, the volume of the extract may be a problem. Several liters of extract cannot easily be passed onto any analytical system. We therefore include a low-pressure liquid chromatography step that efficiently concentrates and initially fractionates the peptides in the extract.

We use wide-bore siliconized glass columns (3.5 × 15 cm) packed with 40- to 63-μm Techoprep C18 (HPLC Technology, Macclesfield, Cheshire, England). The columns are washed with methanol and several bed volumes of water containing 0.1% TFA (No. 8262, Merck). From a suitable reservoir the sample (acidic extract as above, up to 600 ml) may be loaded onto the column by gravity. Overloading may be revealed by addition to the extract of trace amounts of a radioactively labeled peptide having a related chemical structure (e.g., somatostatin 14). If significant amounts of radioactivity appear in the break-through fractions, the column is probably overloaded.

Subsequently, the column is eluted with a gradient of some organic solvent. We routinely use a gradient of acetonitrile in water (plus TFA, 0.1%) from 0 to 80% over 3 hr at a flow rate of 120 ml/hr. Small samples of the eluted fractions are dried down in a vacuum centrifuge, reconstituted in radioimmunoassay buffer, and assayed (or assayed after appropriate dilution).

Immunoreactive material identified by Techoprep chromatography is pooled and applied to gel filtration columns as above (after removal of most of the organic solvent by evaporation) for size fractionation. Resulting immunoreactive peaks are pooled and subjected to analytical HPLC chromatography. By choosing increasingly selective elution conditions (increasingly shallow gradients) and using combined monitoring of the UV absorption and the radioimmunoreactivity, one may gradually increase the purity of the immunoreactive substance and eventually reach a sufficient purity for amino acid, mass, or sequence analysis.

Identification of Isolated Immunoreactive Material

A gross estimation of the molecular size of the isolated material will be apparent from the gel filtration step. It may be that comparisons by analytical HPLC with synthetic model peptides have already pointed to the precise chemical nature of the material. In most cases, however, it may be helpful to determine the mass of the material. This can be performed by mass spectrometry (38), which with modern equipment may be successfully accomplished with amounts of peptide as low as 100–200 pmol. On several occasions we have used plasma desorption mass spectrometry (39). It should be emphasized that mass determination by mass spectrometry requires the same degree of sample purity as amino acid and sequence analysis.

A complete structural characterization can be obtained if one combines a determination of mass with amino acid composition (determined by amino acid analysis) and a structural knowledge of region of the peptide recognized by the monitoring radioimmunoassay and compares these data with the sequence of the propeptide. However, it may be desirable to determine the sequence directly, which may be done with as little as 100 pmol of a small peptide using automated techniques (40). Also, in this instance, a mass determination may be helpful, because the mass data must correspond with the mass calculated from the sequence. Any discrepancy will indicate that the sequence determination was incomplete.

-carboxy groups), whereas the mass of an amidated peptide is not changed (39).

Isolation of Prosomatostatin-Derived Peptide

The peptide prosomatostatin 1–64 from porcine pancreas and gut is isolated and characterized in the following way (13).

1. A peptide corresponding to proSS 20–36 (a gift of P. Andrew, Purdue University) is custom synthesized at Peninsula Laboratories (Belmont, CA).

2. ProSS 20–36 is coupled to bovine serum albumin (Sigma, No. A-4504) using the carbodiimide method described above. Rabbits are immunized as above. After 7 injections, an antiserum (No. 2098) is harvested from one of the rabbits.

3. ProSS 20–36 is iodinated according to the stoichiometric chloramine-T method and the tracer purified by HPLC as described above. The best tracer is eluted at an acetonitrile concentration of 38%.

4. In the resulting radioimmunoassay the final dilution of the antiserum is 1: 20,000. The detection limit is below 5 pmol/liter. In the tabletop version the incubation time is 3 hr, but the sensitivity is around 15 pmol/liter.

5. Pancreas pieces (total 450 g) or ileum mucosa (850 g) are removed from anesthetized pigs and immediately frozen in dry ice (this procedure minimizes nonspecific proteolysis of the precursor products). The tissue is extracted using the acid-ethanol technique described above with ether precipitation of the proteins. The proteins are dissolved in water containing 1 mol/liter urea. Approximately 100 g of tissue is extracted and processed at a time. This amount results in 400 ml of extract.

6. Extract, 400 ml at a time, is applied to a Techoprep column and eluted with a gradient of 10–90% ethanol in water additionally containing TFA. Eluted fractions are diluted 1:50 and assayed for proSS 20–36 immunoreactivity. Immunoreactive fractions (Fig. 2A) are pooled and applied to 50 × 1000 mm gel filtration columns packed with Sephadex G-50 fine grade (Pharmacia) equilibrated and eluted with 0.5 mol/liter acetic acid at a flow rate of 1 ml/min at 4°C (Fig. 2B). Eluted fractions are diluted 1/50 with assay buffer and assayed for proSS 20–36 immunoreactivity.

Fig. 2 Isolation of prosomatostatin 1–64 from ileal mucosa. (A) Elution of immunoreactive material (left ordinate scale) by an ethanol gradient (right ordinate scale) from a Techoprep column. (B) Immunoreactive fractions were pooled, evaporated, and subjected to gel filtration on Sephadex G-50 in acetic acid. (C) Immunoreactive fractions were applied to a Nucleosil C18 column and eluted with a steep gradient of acetonitrile (ACN, for right ordinate scale) in water/trifluoroacetic acid; the eluate was monitored at 226 nm [the absorbance (ABS, on the right ordinate scale) is in arbitrary units]; the immunoreactivity (left ordinate scale) is shown in histogram form. (D) Immunoreactive peak fractions were pooled and applied to a C4 Nucleosil cartridge and eluted with a more shallow gradient (see text). (E) Immunoreactive fractions were then run isocratically on the same column at 33.6% acetonitrile in water/ trifluoroacetic acid. (F) The final isocratic run was performed again on a C18 stationary phase at 37.6% acetonitrile in water/trifluoroacetic acid. For D-F, scales are as in